Splitting MP3 Release into Individual Tracks Options

[GuiltySpark]

I have a digital copy of a CD.

Instead of individual tracks, there are simply two MP3 files (one for each) disc and two CUE files (again, one for each disc).

Infact, this is the contents of the folder:

00-va-big_tunes_ignition-2cd-2009-ute.m3u
00-va-big_tunes_ignition-2cd-2009-ute.nfo
00-va-big_tunes_ignition-2cd-2009-ute.sfv
01-va-big_tunes_ignition__cd_01-ute.cue
01-va-big_tunes_ignition__cd_01-ute.mp3
02-va-big_tunes_ignition__cd_02-ute.cue
02-va-big_tunes_ignition__cd_02-ute.mp3

I understand that there are multiple tools available that would allow me to split the two large files into individual tracks such as:

mp3Directcut
musiCutter
pcutmp3
CUE Splitter
CUETools

Which application is the best/gives the best results?

Cheers guys.

[shearerkazaa]

I'd go with mp3directCut since it's free, is well-updated and has been around for some time already.
No quality loss during the cutting process.

[Soap]

Buy the frickin' CD and stop asking for help with your frickin' scene releases.

[Akkurat]

Yes, BUY THE CD!! You can shove your pirated music, umm, you know where. Sad that there's no TOS rule for these lowlifes.

[2Bdecided]

Sad that there's no TOS rule for these lowlifes.

Rule 9 is often interpreted to mean "don't talk about illegal files", though that's not what it says.

[GuiltySpark]

Sorry, didn't mean to come across as a dirty pirate.

I was rather hoping for an answer or a few pointers though. Can anyone help?

[sizetwo]

Yes. you have already been helped: Go buy the cd.

[AtaqueEG]

Sorry, didn't mean to come across as a dirty pirate.

I was rather hoping for an answer or a few pointers though. Can anyone help?

Next time try not to be so obvious about it.

Not judging, but...

[2Bdecided]

Too lazy to buy the CD, and too lazy to try some free software out for yourself?

I can't be hypocritical. I respect laziness. Try the third one in your list. If the mp3 has sample accurate information, and the cue sheet is correct, you should get a sample-accurate split. If not, it might still be good enough. There's also a thread for pcutmp3 right here on HA, and the output plays gaplessly (which I presume you want) in fb2k.

Cheers,
David.

[chrizoo]

mp3 has sample accurate information

does anyone know what that means ?

[2Bdecided]

mp3 has sample accurate information

does anyone know what that means ?

<ENC_DELAY> : 576
<ENC_PADDING> : 1728
<MP3_ACCURATE_LENGTH> : yes

or similar - look in fb2k properties for the specific mp3 file.

Cheers,
David.

[JunkieXL]

When I need to split an mp3...especially one where the original mp3 was a LAME VBR file... use mp3Directcut and then use the "fix header" feature in foobar.

I haven't had to do this in a while, but it worked like a charm the last time I did. It recreated LAME headers and the files played gaplessly on my iPod.
PS

EDIT: I believe they have updated the program now, so that it creates LAME headers on it's own. You would need to check on that though.

This post has been edited by JunkieXL: Feb 8 2010, 18:46

[greynol]

I just tried that, the result is not gapless. I recommend using a wave editor instead of your ears, especially since the post that resurrected this discussion asked about sample accurate information.

[Takla]

There is an excellent tool for splitting cue+mp3 http://mp3splt.sourceforge.net/mp3splt_page/home.php

It's free software (in all senses) and available for GNU Linux, *BSD, MacOS X, BeOS, Windows.

Mp3Splt-project is a utility to split mp3 and ogg files selecting a begin and an end time position, without decoding. It's very useful to split large mp3/ogg to make smaller files or to split entire albums to obtain original tracks. If you want to split an album, you can select split points and filenames manually or you can get them automatically from CDDB (internet or a local file) or from .cue files. Supports also automatic silence split, that can be used also to adjust cddb/cue splitpoints. You can extract tracks from Mp3Wrap or AlbumWrap files in few seconds. For mp3 files, both ID3v1 & ID3v2 tags are supported.

I've tried it both on cue+mp3 and albumwrap and it works flawlessly as far as I can tell. The resulting mp3s played back gaplessly (when the originals were gapless) and all metadata was retained.

[greynol]

Like with my previous comment, did you verify this with a wave editor?

[JunkieXL]

If I don't hear a gap when listening, then I'm not overly concerned what a wave editor says. I can correct my post to say audibly gapless if you prefer.
PS

This post has been edited by JunkieXL: Feb 8 2010, 20:52

[Takla]

Like with my previous comment, did you verify this with a wave editor?

Not until you mentioned it because I used my ears. Admittedly they are middle aged ears and even when new were not objective measuring instruments but I can tell when the sound actually stops.

I have since put the first two tracks, split from a cue+mp3 using mp3splt, into audacity and they are indeed gapless.

Screens: track one ends at 00:06:29, in the second image this line is marked for clarity:





I'm not trying to promote or advocate use of this software in preference to any other, I posted because I've used it without any difficulty or problems and it was clearly exactly the solution to the question posed. It's multi-platform, free and open source so anyone who cares to can test it in any way they like and perhaps even help improve it, if necessary.

I've recently done a time consuming and thorough clean up of my music collection. I'd forgotten how many dumb ways of doing things I'd been tempted by in the past, such as albumwraps or cue+lossy images, or platform/application specific codecs, formats, or containers, or silly tagging/naming schemes etc etc. I was pleased to find various free tools that could help me get things consistent.

[greynol]

I just tested mp3splt and it's not gapless either.

When I meant a wave editor I meant checking the exact number of samples and looking at the boundary to make sure it is actually seamless which will involve zooming in much closer than that.

Seems like pcutmp3 is the only one so far that will do this correctly.

This post has been edited by greynol: Feb 9 2010, 00:46

[Takla]

I just tested mp3splt and it's not gapless either.

When I meant a wave editor I meant checking the exact number of samples and looking at the boundary to make sure it is actually seamless which will involve zooming in much closer than that.

Seems like pcutmp3 is the only one so far that will do this correctly.

I still didn't figure out how to listen with anything except my my ears, so I'll go with what an earlier poster said, which is it's audibly gapless, or perhaps say "there are no audible gaps, and the listening experience is seamless and non different than listening to the original cue+file". If there is a gap, but it is beyond the ability of the human ear to detect it, this is where I stop caring. If this was a tool for lossless audio, and absolute fidelity to the original was intended, then I might have a concern, but this is mp3 so the question of an inaudible gap seems to me to be about as important as the infra sound and ultra sound harmonics which the golden ears brigade feel are lost by compression. If someone can actually abx a gap that isn't even easily apparent in a wav editor then I'll be surprised, also rather impressed....and actually awe struck if it's abx'd by someone who doesn't know exactly where the gap is supposed to be, or even that it exists.

[greynol]

That's a lot of posturing for something that's quite easy to ABX:

CODE

foo_abx 1.3.1 report
foobar2000 v0.9.4.2
2010/02/08 16:34:13

File A: C:\clip1.wav
File B: C:\clip2.wav

16:34:13 : Test started.
16:34:29 : 01/01 50.0%
16:34:33 : 02/02 25.0%
16:34:36 : 03/03 12.5%
16:34:43 : 04/04 6.3%
16:34:46 : 05/05 3.1%
16:34:52 : 06/06 1.6%
16:34:55 : 07/07 0.8%
16:34:59 : 08/08 0.4%
16:35:03 : 09/09 0.2%
16:35:09 : 10/10 0.1%
16:35:18 : 11/11 0.0%
16:35:22 : Test finished.

----------
Total: 11/11 (0.0%)

Would you like me to post my clips?

EDIT: This was done by cutting a piece of gapless audio into two using mp3splt, decoding the two pieces to wave with foobar2000 and putting them back together in a wave editor and cropping down to a 10 second sample (clip2.wav). clip1.wav is the exact same section of audio from the decoded mp3 before splitting. Yes, the gaps created using mp3DirectCut are with this sample are audible as well!

I can make samples using pcutmp3 to show you guys how this is done correctly if you like.

This post has been edited by greynol: Feb 9 2010, 02:07

[Takla]

I don't mind if you post your clips or not. When I've used the tool I couldn't hear a gap, neither when playing the mp3s nor when playing concatenated wavs. Nor can I see even a suggestion of any gap in a wav editor. I have subsequently zoomed in so I'm looking at tenths, then hundredths, then thousandths(!) of seconds and still there is no gap apparent, so I'm really puzzled at the clear difference in your results. The mp3s are unremarkable joint stereo VBR. I will have to create some cue+mp3s with different tools and and also use various splitters and check the differences, if any. I was (am) perfectly satisfied with the files produced by mp3splt so I no longer have the original cue+mp3s or albumwraps. But it's past 1 am here so I won't be doing it immediately.

[greynol]

thousandths(!)

Not sure what's with the exclamation point. Gaps that are "mere" thousandths of a second can easily be audible. The gap in clip2 is about 6ms wide.

 clip1.flac ( 915.51K ) Number of downloads: 70
 clip2.flac ( 912.77K ) Number of downloads: 79


This post has been edited by greynol: Feb 9 2010, 02:50

[aclo]

thousandths(!)

Not sure what's with the exclamation point. Gaps that are "mere" thousandths of a second can easily be audible. The gap in clip2 is about 6ms wide.

 clip1.flac ( 915.51K ) Number of downloads: 70
 clip2.flac ( 912.77K ) Number of downloads: 79

And even without trying this, it is not that hard to believe: clicks from LPs are of the order of 1ms from what I remember, yet clearly audible. I do realize that spikes are probably more easily audible than lack of sound of the same duration, but this just demonstrates that features that short aren't obviously inaudible.

Now a temporally short spike contains loads of frequencies (because short so that its FT extends to high frequencies) so maybe this helps it being more audible than one would expect. But then again a blank of the same duration presumably corresponds to adding to the signal exactly the right spike to cancel it (ie, adding the signal but with its phase shifted everywhere by pi). But I am sure that the noticeability of the presence and absence of certain frequencies are not simply related . So...

But it's 3 in the morning so maybe I just need some sleep.

[gfxnow]

Ok so lets get this straight, pcutmp3 is the only splitter/cutter that's doing it properly? How about mp3directcut 2.11? Does that mess up the number of samples and not retain encoder delay and padding info for gapless playback?

[greynol]

Ok so lets get this straight, pcutmp3 is the only splitter/cutter that's doing it properly?

I have not tested CUETools, but amongst the other contenders listed in the OP, pcutmp3 is the only one that can cut without gaps.

How about mp3directcut 2.11? Does that mess up the number of samples and not retain encoder delay and padding info for gapless playback?

mp3DirectCut 2.11 does not alter the encoder delay and padding values in order to achieve gapless playback. Preserving the bit reservoir also needs to be taken into account in order to achieve gapless playback. I would be surprised if mp3DirectCut does this either.

This post has been edited by greynol: Feb 9 2010, 04:34